Prediction of Lateritic Soils UCS Using Index Properties

Five predictive models were developed and calibrated using both regression and soft computing techniques: Gaussian process regression (GPR), particle swarm optimization–artificial neural network (PSO-ANN), support vector machine (SVM), gene expression programming (GEP), and multivariate nonlinear regression (MNLR) to predict UCS using index properties. The model’s performance was evaluated using statistical indices such as coefficient of determination (R²), root mean squared error (RMSE), mean absolute error (MAE), and mean absolute percentage error (MAPE) and residual standard error (RSE).